206 THE VOYAGE OF H.M.S. CHALLENGER. 



furrows, separating the batteries of nematocysts. The gland cells of each group are 

 separated from one another by slender thread-like cells, similar to those between and 

 beneath the nematocysts. A nervous layer is more or less well marked at the base of 

 the ectoderm, but is granular in the specimens examined. An ectodermal row of 

 muscular fibres occurs, which is in some parts more fully developed than in AntipatheS 

 and Antipathella. The ectoderm is about 0'05 mm. thick near the middle of a tentacle, 

 but towards the base the layer becomes much thickened, and the nematocysts are 

 observable as small isolated clusters of cells, which extend for no great depth into the 

 general mass of cells. In sagittal sections of a zooid (PL XV. fig. 4) the batteries of 

 nematocysts are often indistinguishable on the surface of the body-wall and peristome, 

 where the ectoderm is much thickened and considerabfy modified. In such situations 

 the greater portion of the ectoderm consists of an irregular faintly-stained reticulum, 

 enclosing hyaline cells. Nearly all the histological elements shown in the tentacles 

 (PI. XV. fig. 3, ect) have either disappeared or become so much modified as to be no 

 longer recognisable. In horizontal sections the ectoderm, under such circumstances, is 

 seen to be invaded by a number of slender mesoglceal processes, which are often branched, 

 and undoubtedly form a part of the reticulum referred to. I have been unable to decide 

 how far these processes extend, but the whole structure is so remarkable as to require 

 a renewed study. The appearance presented recalls the condition of the ectodermal 

 surface of the mesoglcea in certain Challenger Actiniaria described by R. Hertwig {e.g., 

 Ilyanthopsis, Hormathia, and Phellia spinifera)} In Ilyanthopsis longifilis the 

 muscular pleats of the oral disc are slightly arborescent and arranged close together. At 

 the free edge of each pleat mesoglceal fibres radiate into the ectoderm and are for some 

 distance connected into bundles. In Phellia spinifera the mesoglceal ingrowths are 

 stronger, and bear muscular fibres on each side throughout their entire length, whilst in 

 some parts they become fused together into an irregular reticulum. It appears probable 

 that the more complex muscular folds of many Hexactiniae are due to similar ingrowths 

 of the mesoglcea. It is worthy of note that in Leiopathes the longitudinal muscular 

 fibres of the ectoderm are not confined to the normal surface of the mesoglcea, but may 

 be seen to follow the outline of the mesoglceal processes for a short distance as in 

 Ilyanthopsis. This layer is therefore not always flat as in the genera previously described. 

 On this account there is a closer resemblance to the ectodermal muscular layer of 

 Cirripathes propinqna, but the fibres in Leiopathes are not so thick. 



Stomodamm. — The stomodaeum of Leiopathes glaberrinia is rather short, and does 

 not extend into the lower two-thirds of the ccelenteron, excepting in that portion of it 

 which borders the transverse axis. In this region the stomodasum is continued along 

 the free margin of the transverse mesenteries, and the cells forming the rounded free 

 extremity of each mesenterial filament are apparently derived from it (PI. XV. fig. 4). 



1 Zool. Chall. Exp., pt. lxxiii. pi. ii. figs. 2, 3, and 8. 



